Sie sind auf Seite 1von 6

Journal of Colloid and Interface Science 227, 1–6 (2000)

doi:10.1006/jcis.2000.6810, available online at on

Surface Pressure, Hysteresis, Interfacial Tension, and CMC of Four

Sorbitan Monoesters at Water–Air, Water–Hexane,
and Hexane–Air Interfaces
Leena Johanna Peltonen1 and Jouko Yliruusi
Department of Pharmacy, Pharmaceutical Technology Division, University of Helsinki, Helsinki FIN-00014, Finland

Received March 2, 1999; accepted March 6, 2000

In this study, the behavior of sorbitol surfactants with differ-

The purpose of this study was to investigate the interfacial prop- ent alkyl chain lengths or structures was analyzed. Four sorbitan
erties of sorbitan monoesters (Span 20, 40, 60, and 80). The sur- monoesters (Span 20, Span 40, Span 60, sorbitan monostearate,
face pressure was investigated at the water–air interface using a
and Span 80) which differ from each other by the length (Span
Langmuir–Blodgett apparatus. Interfacial tensions at n-hexane–
air and water–n–hexane interfaces were measured by a du Nouy
20, 40, and 60) or the structure (Span 60 and 80) of the hydrocar-
tensiometer. The effects of different surface-active agents and their bon chain were tested. Dynamic surface pressure and hysteresis
concentrations on the interfacial properties of surfactant films were of the spread films of sorbitol surfactants were measured at the
determined. With saturated sorbitan monoesters the lengthening of water–air interface by the Langmuir–Blodgett technique. CMCs
the hydrocarbon chain increases the collapse pressure and molecu- of the surface-active agents in the water–n-hexane system were
lar area at the water–air interface. Unsaturated Span 80 had a lower determined with the aid of interfacial tension measurements us-
collapse pressure and a larger molecular area than its saturated ing a du Nouy tensiometer. Also interfacial tensions of surface-
counterpart Span 60. Under compression–expansion cycles, all sor- active agents in n-hexane against air were measured.
bitan monoesters showed hysteresis effects. At the n-hexane–air in-
terface there were no differences in the interfacial tension between MATERIALS AND METHODS
different sorbitan monoesters. At the water–n-hexane interface, dif-
ferences in CMCs were small, but the surface excess of Span 80 was Materials
markedly smaller and the molecular area larger than the corre-
sponding values of other sorbitan monoesters. °C 2000 Academic Press All chemicals were used as received. The surface-active
Key Words: interfacial tension; surface pressure; CMC; liquid– agents Span 20, Span 40, Span 60, and Span 80 were all pur-
liquid interface; n-hexane; sorbitan monoesters. chased from Fluka Chemie AG, Switzerland. Some physical
properties of sorbitan monoesters are summarized in Table 1.
The structure of sorbitan monoesters is presented in Fig. 1.
INTRODUCTION Hexane (purity ≥99%, Fluka Chemie AG, Switzerland) was
used as a spreading agent in surface pressure measurements and
Despite the wide use of sorbitan esters, rather few studies have also as the oil phase in oil–air and water–oil measurements. The
been conducted on their interfacial tension and surface pressure. water used was ultrapure Millipore water.
Interfacial film strength is an important contributor to emulsion
stability, and there are some articles on the evaluation of emul- Surface Pressure
sion stability by surface tension measurements (1, 2). To our
The surface pressure was measured by a computer-
knowledge, all the studies of sorbitan esters concern only a few
controlled Langmuir–Blodgett instrument (KSV Minitrough,
members of the sorbitan ester family. Span 20, sorbitan mono-
KSV, Finland; Fig. 2). The measurement itself was made by the
laurate (3–6), and Span 80, sorbitan monooleate (4–10), are the
Wilhelmy plate method, placing the plate partly under the sur-
most widely used, but also Span 40, sorbitan monopalmitate
face of the subphase and recording the force with which the plate
(7), and Span 83, sorbitan sesquioleate, or Span 85, sorbitan tri-
was pulled downward by the surface film. This force was mea-
oleate, have been used a few times (5–7). There is no article on
sured using electrobalance. By adjusting movable barriers, the
systematic research on the surface and interfacial properties of
area available to the film could be modified. When the film area
sorbitan esters.
diminished molecules lying flat on the liquid surface were forced
1 To whom correspondence should be addressed at the Department of
into more vertical positions where they required less space. The
Pharmacy, Pharmaceutical Technology Division University of Helsinki, P.O. ease of compressing the film and the area per molecule at a given
Box 56 (Viikinkaari 5), FIN-00014 University of Helsinki, Finland. E-mail: pressure represent how readily a molecule tolerates the presence of its neighbors.
1 0021-9797/00 $35.00
Copyright °C 2000 by Academic Press
All rights of reproduction in any form reserved.

Some Physical Parameters of Sorbitan Monoesters

Span 20 Span 40 Span 60 Span 80

HLB 8.6 6.7 4.7 4.3

Mw (g/mol) 346 403 431 429
δo (cal/cm3 )1/2 13.52 12.58 11.59 11.38
Cw /Co 85 0.43 1.7 × 10−3 5.5 × 10−4

Note. (All values are from the literature): HLB (hydrophilic–lipophilic

balance, from Ref. 12), Mw (molecular weight, 11), δo (solubility parameter,
FIG. 1. The structure of sorbitan monoesters. R is –OOC(C11 H23 ) for 13, 14), Cw /Co (water–oil partition coefficient, 13, 14).
Span 20, –OOC(C15 H31 ) for Span 40, –OOC(C17 H35 ) for Span 60, and
–OOC(C17 H33 ) for Span 80 (12).
film was deposited. As a control test, a blank compression with-
out a film was measured. During this measurement, which took
The trough was made out of a solid piece of PTFE and the 50 min, the pressure developed did not exceed 0.15 mN/m. The
barriers of hydrophilic material, Delrin, which ensured that the dynamic values were measured continuously for at least 50 min
film did not go under the barrier. The barrier position was con- without disturbing the interface. The hysteresis measurements
trolled by a servo-controlled DC motor. The motor moved the took approximately 9 h.
barrier holder using a tooth belt. The position of the barrier was An ethanol–hexane mixture (1 : 9 v/v) in which surface-active
measured using an optical encoder. The trough was glued to an agents were dissolved was used as a spreading solution. This
aluminum base plate, in which there were water channels for spreading solution (20 to 50 µl) was then deposited on the water–
thermostation of the subphase. Thermostation was performed air interface. The measurement was started 10 min later to en-
using a circulator connected to the trough. sure that all the ethanol–hexane solvent had vaporized from the
Before each measurement, the trough and the barriers were surface. In order to ensure that 10 min waiting time was long
washed with pure ethanol and rinsed carefully with ultrapure enough, a control test was run out: without starting the measure-
Millipore water. This washing and rinsing was done twice. Also ment the differences in surface pressure were monitored for 2 h
all the glassware used was cleaned twice with ethanol and then after the film was deposited on the water–air interface. All the
rinsed several times with Millipore water before use. The barriers measurements were run five times and the results were calcu-
were moved to the ends of the trough manually, and the trough lated as the mean of these five different measurements. With all
was filled with Millipore water (45 ml). The barriers were driven the sorbitan monoesters the standard deviation in surface pres-
manually to the center, and the water surface between the bar- sure measurements was less than 0.5 mN/m.
riers was cleaned with a suction pipette. After that the barriers The dimensions of the trough were such that the total surface
were driven to the starting position and the Wilhelmy plate was area of 6930 mm2 (231 × 30 mm) was available for compres-
settled so that one-third of the plate was under the water surface. sion. The measurement rate (rate of the barriers) was 2 mm/min
In that position the surface pressure value was zeroed and the (=2 mm2 /s).

FIG. 2. The block diagram of the Langmuir–Blodgett surface balance measurement station. The Wilhelmy plate is connected to the microelectronic feedback
system for surface pressure control. The force was measured using the electrobalance, but there is also a microelectronic feedback system for surface pressure
control (modified from Ref. 15).

Before the film balance work with the chosen surface-active

agents the technique was checked against a pressure–area
isotherm found in the literature: stearic acid (16). Occasionally
throughout this work, the validity of the technique was checked
with this standard. The complete spreading of film-forming ma-
terial was ensured by changing the spreading solution (chloro-
form, hexane, hexane–ethanol (9 : 1)) and the depositing method
of the film onto the water surface. All these above-mentioned
measurements were reproducible, which is believed to prove that
the film-forming material was spread completely on the water

Interfacial Tension
The interfacial tension was measured with a du Nouy ten- FIG. 3. The surface pressure–molecular area curves of sorbitan monoesters
siometer (KSV Sigma 70, KSV, Finland) with a platinum ring. (T = 295 K).
The interfacial tension was calculated from the maximum force
needed to separate the ring from the interface.
Several dilutions of each surface-active agent were made in The shape of the Span 40 and 60 curves is also steeper, which
hexane. For the interfacial tension measurements, 10 ml of each implies that these molecules tolerate compression better and are
dilution was needed. The interfacial tension of surface-active more readily associated with each other than Span 20 and 80.
agent in hexane against air was measured first. Five measure- The increase in surface pressure with increasing alkyl chain
ments were made at 5-min intervals from the same sample and length indicates that the compounds derive the surface activity
means and standard deviations were calculated. primarily from the alkyl group. In a monolayer, sorbitan mo-
Second, the interfacial tension between hexane and water was noester molecules are oriented so that each hydrocarbon chain
measured with the surface-active agents dissolved in the hexane lies on the oil side of the interface and each sorbitan ring lies
phase. Water was settled as a lower phase, and the surface-active on the aqueous phase (6). The significant increment in collapse
agent–hexane solution was gently poured onto the water phase. pressure between Span 20 and Span 40 suggests that lengthen-
The system was not stirred to avoid emulsification. The phases ing of the hydrocarbon chain from 12 to 16 carbon atoms dom-
were allowed to stabilize for 30 min before the measurement (the inates the molecular properties in this region. It was observed
surface-active agent in the hexane phase). Again, five different earlier that emulsifier molecules cannot be associated readily
measurements were carried out at 30-min intervals in order to unless the hydrocarbon chains contain at least 8 carbon atoms
ensure a sufficently long stabilization time. Thirty minutes of (6). With 12 carbon atoms (Span 20) the polarity, caused by the
stabilization was determined by measuring the interfacial ten- shorter fatty acid chain length, has been reported to result in
sion as a function of the age of the interface until a constant aqueous phase solubilization, which apparently hinders interfa-
value was reached. cial molecular interactions (5). Diethanolamine derivatives, also
From the results of the interfacial tension measurements nonionic surface-active agents, have been noticed to behave as
CMCs were extrapolated. The instrument was calibrated against sorbitan monoesters: the longer the alkyl chain, the higher the
Millipore water. All the surface pressure and interfacial tension collapse pressure observed and the larger the molecular area
measurements were taken at room temperature (22.0 ± 0.5◦ C). (18).
In comparisons of surface pressure–molecular area curves of
Span 60 (saturated) and 80 (unsaturated), the molecular area
RESULTS of Span 80 is larger than that of Span 60. With Span 80 there
is a discontinuity point at approximately 33 Å2 and 23 mN/m.
Water–Air Interface
Surface areas above this discontinuity point are larger with Span
The shapes of the curve of the dynamic surface pressure of 80 than with Span 60. When stearic and oleic acids are compared
sorbitan monoesters (Fig. 3) at the water–air interface are sim- (Span 60 is a sorbitan ester of stearic acid and Span 80 a sorbitan
ilar to those of alkyldimethylphosphine oxides (also nonionic ester of oleic acid), the surface pressure–molecular area curves
surface-active agents (17)). The collapse pressure of saturated behave in the same way, the molecular area of stearic acid being
sorbitan monoesters is lowered by shortening the hydrocarbon smaller at the same surface pressure compared to that of oleic
chain. This lowering is more obvious between 12 (Span 20) and acid (19).
16 (Span 40) carbon atoms than between 16 and 18 (Span 60) When measurements are made after 4 h of stabilization instead
carbon atoms. Also the collapse pressure is more determinable of 10 min, the discontinuity point is more obvious with Span 20
with Span 40 and 60, while with Span 20 and 80 the collapse and 80 (Fig. 4), though the shift to the left is not so dramatic
point is not so clear. with Span 20. With Span 80 the discontinuity point after 4 h

FIG. 4. The surface pressure–molecular area curves of sorbitan monoesters, FIG. 6. The hysteresis of Span 40: the numerals are the numbers of the
the effect of waiting time before starting the measurement: the dashed line is 4 h compression–expansion cycle.
before measurement and the solid line is 10 min before measurement.

continuity point of Span 20 did not fall as the hysteresis ad-

of stabilization is approximately 23 Å2 and 30 mN/m. It was vanced. With Span 80 the shape of the curve changed, and the
noticed earlier that Span 80 has a tendency to form a stronger drop of the surface pressure was more obvious as the hystere-
film given sufficient time (4). With Span 40 and 60 a small shift sis advanced. This is partly caused by the great excess of Span
to the left can also be seen, but the shape of the curve is similar 80 which was injected to the interface in order to have hys-
(Fig. 4). teresis in the same molecular area as all the other sorbitan mo-
noesters. That caused the surface pressure during the first com-
pression to be consistently above the discontinuity point (Figs. 3
All the sorbitan monoesters showed hysteresis effects when and 8).
the monolayers were subjected to compression–expansion One reason for the strong drop in surface pressure in the be-
cycles (Figs. 5–8). Between each following compression– ginning of expansion may be the molecular orientation, e.g.,
expansion cycle there was a shift to the left. This means that flipping of hydrocarbon tails from a horizontal to a vertical po-
material is lost from the surface during compression. However, sition. Relaxation may also be partly due to the redistribution
that material is not respread on the surface during the expansion. of unequally distributed molecules resulting from the disturbing
This loss of material is probably due to the fact that sorbitan mo- effects of compression (21).
noesters, though practically insoluble in water, are dispersible The area of hysteresis decreased markedly between the
in the water phase during the compression process (20). successive compression curves, the difference between the first
In the beginning of the expansion the surface pressure falls and the second cycle being the greatest. Similar behavior was
strongly with all sorbitan monoesters, with Span 20 close to reported previously with dipalmitoyl lecithin (22) and stearic
0 mN/m. The collapse pressure of Span 40 and 60 and the dis- acid (21).

FIG. 5. The hysteresis of Span 20: the numerals are the numbers of the FIG. 7. The hysteresis of Span 60: the numerals are the numbers of the
compression–expansion cycle. compression–expansion cycle.

Molecular Area of Sorbitan Monoesters
Span 20 Span 40 Span 60 Span 80
A (A2 ) 40 37 38 46

where R is the gas constant, T absolute temperature, γ interfacial

tension, and c concentration. From the surface excess, the area
per molecule, A, just below the CMC was calculated according

A= , [2]
FIG. 8. The hysteresis of Span 80: the numerals are the numbers of the where N is Avogadro’s number. The molecular area values are
compression–expansion cycle. summarized in Table 2.
Table 2 shows that Span 40 had the smallest molecular
Oil–Air Interface area though the differences between saturated sorbitan mo-
All the sorbitan monoesters were soluble in n-hexane, by noesters were small. Unsaturated Span 80 shows a larger area
slight warming for Span 40 and Span 60, and CMC could not per molecule.
been determined. Sorbitan monoesters, too, have no effect on Generally it has been noticed that for a homologous series
the interfacial tension of n-hexane against air. Similar behavior of nonionic surfactants the increase in the hydrophobic chain
was noticed in earlier studies with mineral oil (5). The surface length increases the apparent CMC at the oil–water interface (4).
tension of mineral oil against air did not change after the addi- However, with our surface-active agents the differences in CMC
tion of the Spans. This phenomenon has been explained by the were quite small (all the CMCs approximately 0.0070 mg/ml);
fact that in the absence of a hydrophilic phase there is no driving in all our research, especially those between 16 and 18 saturated
force for interfacial adsorption and multilayer formation. carbon atoms seemed to be not so dominant. The reason for this
may be the small difference in the hydrophobic chain length,
Water–Oil Interface only two carbon atoms, between these two surface-active agents.

From the results of the interfacial tension measurements at CONCLUSIONS

the water–n-hexane interface (Fig. 9), the surface excess, 0,
was calculated by means of the Gibbs’ adsorption isotherm Interfacial tension and surface pressure techniques can be used
to determine the interactions in the presence and the absence
1 dγ of different surface-active agents (23–38). These measurements
0=− × , [1]
RT d ln c have appeared to be useful, for example, in preformulation stud-
ies on emulsifier films: the stronger the interfacial film, the more
stable the resultant emulsion.
In this study, the effect of the concentration of various sur-
factants (Span 20, 40, 60, and 80) on the interfacial proper-
ties of surfactant films were determined. With saturated sorbitan
monoesters the lengthening of the hydrocarbon chain increases
the collapse pressure and molecular area at the water–air inter-
face, but unsaturation of the hydrocarbon chain lowers the col-
lapse pressure and increases the molecular area. All the sorbitan
monoesters used showed hysteresis effects under compression–
expansion cycles. With these sorbitan monoesters, unsaturation
of the hydrocarbon chain seemed to be more dominant in deter-
mining the interfacial properties at the water–n-hexane interface
than in determining the hydrocarbon chain length. The differ-
ences between the saturated 16 and 18 carbon atoms in the chain
were especially small.
FIG. 9. The effect of concentration of sorbitan monoesters on interfacial Some interfacial properties of sorbitan monoesters were
tension at water–n-hexane interface. The error bars are within the symbols. highlighted, but in this area of research there are still many

phenomena that are not understood or explained as yet. Work 18. Ghaicha, L., Leblanc, R. M., and Chattopadhyay, A. K., J. Phys. Chem. 96,
currently under way in our laboratory aims at further clarifying 10948 (1992).
these properties of sorbitan ester-type surfactant systems. 19. Feher, A. I., Collins, F. D., and Healy, T. W., Aust. J. Chem. 30, 511
20. Reynolds, J. E. F. (Ed.), “Martindale, The Extra Pharmacopoeia” 30th ed.
REFERENCES p. 1032, Pharmaceutical Press, London, 1993.
21. Munden, J. W., Blois, D. W, and Swarbrick, J., J. Pharm. Sci. 58, 1308
1. Ishii, F., Takamura, A., and Ogata, H., J. Pharm. Pharmacol. 40, 89 (1988). (1969).
2. Takamura, A., Ishii, F., Noro, S., Tanifuji, M., and Nakajima, S., J. Pharm. 22. Galdston, M., and Shah, D. O., Biochim. Biophys. Acta 137, 255 (1967).
Sci. 73, 91 (1984). 23. Dahanayake, M., Cohen, A. W., and Rosen, M. J., J. Phys. Chem. 90, 2413
3. Phares, R. E., Jr., J. Pharm. Sci. 54, 408 (1965). (1986).
4. Wan, L. S. C., and Lee, P. F. S., Can. J. Pharm. Sci. 8, 136 (1973). 24. Rosen, M. J., Cohen, A. W., Dahanayake, M., and Hua, X.-Y., J. Phys.
5. Opawale, F. O., and Burgess, D. J., J. Colloid Interface Sci. 197, 142 (1998). Chem. 86, 541 (1982).
6. Boyd, J., Parkinson, C., and Sherman, P., J. Colloid Interface Sci. 41, 359 25. Nagarajan, R., and Wasan, D. T., J. Colloid Interface Sci. 159, 164 (1993).
(1972). 26. Liggieri, L., Ravera, F., and Passerone, A., J. Colloid Interface Sci. 169,
7. Rabiskova, M., Song, J., Opawale, F. O., and Burgess, D. J., J. Pharm. 238 (1995).
Pharmacol. 46, 631 (1994). 27. Göbel, J. G., and Joppien, G. R., J. Colloid Interface Sci. 191, 30 (1997).
8. Kawashima, Y., Hino, T., Takeuchi, H., and Niwa, T., Chem. Pharm. Bull. 28. Schott, H., J. Pharm. Sci. 58, 1521 (1969).
40, 1240 (1992). 29. Corkill, J. M., Goodman, J. F., and Ottewill, R. H., Trans. Faraday Soc. 57,
9. Campanelli, J. R., and Wang, X., J. Colloid Interface Sci. 190, 491 (1997). 1627 (1961).
10. Criddle, D. W., and Meader, A. L., Jr., J. Appl. Phys. 26, 838 (1955). 30. Corkill, J. M., Goodman, J. F., and Harrold, S. P., Trans. Faraday Soc. 60,
11. Budavari, S., O’Neil, M. J., Smith, A., Heckelman, P. E., and Kinneary, J. F. 202 (1964).
(Eds.), “The Merck Index,” 12th ed. p. 1490, Merck, Whitehouse Station, 31. Silvestri, S., Ganguly, N., and Tabibi, E., Pharm. Res. 9, 1347 (1992).
NJ, 1996. 32. Yamaguchi, T., Nishizaki, K., Itai S., Hayashi, H., and Ohshima, H., Pharm.
12. Florence, A. T., and Attwood, D., “Physicochemical Principles of Phar- Res. 12, 1273 (1995).
macy,” 3rd ed. Macmillan Press Ltd., Easton, Bristol, 1998. 33. Varadaraj, R., Bock, J., Valint, P., Jr., Zushma, S., and Thomas, R., J. Phys.
13. Davies, J. T., and Rideal, E. K., “Interfacial Phenomena,” Academic Press, Chem. 95, 1671 (1991).
New York, NY, 1961. 34. Elworthy, P. H., and Florence, A. T., J. Pharm. Pharmacol. 21, 72 (1969).
14. Schott, H., J. Pharm. Sci. 84, 1215 (1995). 35. Elworthy, P. H., and Mysels, K. J., J. Colloid Interface Sci. 21, 331 (1966).
15. “KSV Minitrough Instruction Manual,” Finland, 1993. 36. Yoon, K. A., and Burgess, D. J., Pharm. Res. 13, 433 (1996).
16. Cook, H. D., and Ries, H. E., Jr., J. Phys. Chem. 60, 1533 (1956). 37. Yoon, K. A., and Burgess, D. J., J. Pharm. Pharmacol. 49, 478 (1997).
17. Ferrari, M., Liggieri, L., Ravera, F., Amodio, C., and Miller, R., J. Colloid 38. Mollet, C., Touhami, Y., and Hornof, V., J. Colloid Interface Sci. 178, 523
Interface Sci. 186, 40 (1997). (1996).